Control method for participation of user side energy storage in virtual power plant operation

Abstract

The invention relates to the technical field of power grid control, and discloses a control method for participation of user side energy storage in virtual power plant operation, and the method comprises the steps: obtaining operation data, and determining a current value of a virtual state charge representing the comprehensive energy storage capability, an economic driving factor and a power grid driving factor; setting an event memory factor to record a historical power grid event, and dynamically adjusting the operation boundary of the virtual state charge based on the event memory factor; selecting a corresponding mode according to whether the current state is in a power grid emergency event or not, and calculating a target value of a virtual state charge in combination with the driving factor and a dynamically adjusted operation boundary; and finally, calculating and issuing a final power instruction based on the deviation between the current value and the target value of the virtual state charge. According to the method, when the user side energy storage participates in the operation of the virtual power plant, the operation economy and the power grid support reliability are effectively considered through adaptive control, and the adaptability and robustness of the control method are obviously improved.

Description

Technical Field

[0001] This invention relates to the field of power grid control technology, specifically to a control method for user-side energy storage to participate in the operation of a virtual power plant. Background Technology

[0002] Driven by the goal of achieving "dual carbon" emissions, the penetration rate of renewable energy sources, represented by wind power and photovoltaics, in the power grid has rapidly increased. However, their inherent intermittency and volatility pose serious challenges to the stable operation of the power system. Virtual power plants, as an advanced energy management model, have emerged to address this challenge. They aggregate distributed power sources, controllable loads, and energy storage resources scattered throughout the power grid through information technology, forming a unified, coordinated, and flexibly dispatchable whole. Among these, user-side energy storage, due to its flexible deployment and rapid response, has become an indispensable key regulation resource in virtual power plants, playing a crucial role in enhancing the grid's capacity to absorb clean energy.

[0003] Currently, in practical applications of user-side energy storage participating in the operation of virtual power plants, the control strategies mostly revolve around two core objectives: first, to perform peak shaving and valley filling based on time-of-use pricing or market signals to obtain economic benefits; and second, to provide temporary frequency regulation or peak shaving ancillary services in response to signals such as frequency deviation when receiving grid dispatch instructions. These traditional control methods typically map economic or grid signals directly into energy storage charging and discharging power commands within pre-defined, fixed physical operating boundaries to achieve dispatch management of energy storage units.

[0004] However, the inventors discovered significant shortcomings in the aforementioned control strategy during their research. First, the use of fixed operating boundaries makes the control strategy lack adaptability to long-term grid operational risks. When the grid experiences frequent emergencies due to extreme weather or other reasons, traditional strategies struggle to remember these historical conditions and proactively reduce operating ranges to reserve more backup capacity, resulting in insufficient response capabilities in subsequent emergencies. Second, existing strategies often fail to effectively balance economic objectives with grid safety objectives. Overemphasizing economic efficiency may lead to energy storage being under-powered when the grid truly needs support. Furthermore, the lack of unified comprehensive state indicators to guide control complicates the control logic under multi-objective coordination.

[0005] Therefore, this invention proposes a control method for user-side energy storage to participate in the operation of a virtual power plant, in order to address the shortcomings of existing technologies. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention provides a control method for user-side energy storage to participate in the operation of a virtual power plant, which solves the problems of poor adaptability of existing control strategies and difficulty in effectively balancing operational economy and grid support reliability.

[0007] To achieve the above objectives, the present invention provides a control method for user-side energy storage to participate in the operation of a virtual power plant, comprising:

[0008] S1. Obtain real-time operating data and external environment data of user-side energy storage;

[0009] S2. Based on the real-time operating data and the external environment data, determine the economic driving factor and the power grid driving factor;

[0010] S3. Determine the current value of the virtual state charge used to characterize the comprehensive economic and grid support capabilities of the user-side energy storage;

[0011] S4. Set an event memory factor to record historical power grid emergency events, and dynamically adjust the operating boundary of the virtual state charge based on the event memory factor;

[0012] S5. Based on the economic driving factor and the power grid driving factor, and in conjunction with the operating boundary of the dynamically adjusted virtual state charge, calculate the target value of the virtual state charge.

[0013] S6. Based on the deviation between the target value of the virtual state charge and the current value of the virtual state charge, the expected charging and discharging power is calculated.

[0014] S7. Based on the physical constraints of the user-side energy storage, the desired charging and discharging power is modified to generate a final power command for responding to the virtual power plant dispatch and then issued for execution.

[0015] Preferably, in step S1, the real-time operating data includes the physical state of charge of user-side energy storage; the external environment data includes real-time market electricity prices and grid frequency.

[0016] Step S1 further includes: attaching a unified timestamp calibrated by the Network Time Protocol to the real-time running data and the external environment data from different sources, and performing validity verification on the real-time running data and the external environment data.

[0017] Preferably, in step S2, the step of determining the economic driving factor and the power grid driving factor based on the real-time operating data and external environment data includes:

[0018] The economic driving factor is calculated and determined based on the difference between the real-time market electricity price and the preset reference electricity price in the external environment data.

[0019] The grid driving factor is calculated and determined based on the deviation between the grid frequency and the reference frequency in the external environment data, and the rate of change of the deviation between the grid frequency and the reference frequency.

[0020] Preferably, the economic driving factor Feco The formula for calculating (t) is:

[0021] F eco (t)=k eco ·(C base (t)-C real (t));

[0022] In the formula, F eco (t) represents the economic driving factor at the current time t; k eco C is the economic driving factor. base (t) represents the preset reference electricity price at the current time t; C real (t) represents the real-time market electricity price at the current time t;

[0023] The power grid driving factor F grid The formula for calculating (t) is:

[0024]

[0025] In the formula, F grid (t) represents the power grid driving factor at the current time t; k p k is the proportionality coefficient. d is the differential coefficient; Δf(t) is the deviation between the grid frequency and the reference frequency at the current time t; This represents the rate of change of the deviation between the grid frequency and the reference frequency.

[0026] Preferably, in step S3, the step of determining the current value of the virtual state charge used to characterize the integrated economic and grid support capabilities of the user-side energy storage includes:

[0027] At the initial moment of the method execution, the initial physical state of charge of the user-side energy storage is used as the current value of the virtual state charge;

[0028] In subsequent operating cycles, the current value of the virtual state charge is iteratively updated based on the final power command issued and executed in the previous control cycle.

[0029] Preferably, in step S4, the step of setting an event memory factor to record historical power grid emergency events and dynamically adjusting the operating boundary of the virtual state charge based on the event memory factor includes:

[0030] Set an event memory factor to record the frequency or severity of historical power grid emergency events;

[0031] And based on the event memory factor, the operating boundary of the virtual state charge is dynamically adjusted;

[0032] The step of dynamically adjusting the operating boundary of the virtual state charge includes:

[0033] When the frequency or severity of historical power grid emergency events represented by the event memory factor increases, the operating boundary of the virtual state charge is tightened.

[0034] When the frequency or severity of historical power grid emergency events represented by the event memory factor decreases, the operating boundary of the virtual state charge is relaxed.

[0035] Preferably, the update formula for the event memory factor M(t) is:

[0036]

[0037] In the formula, M(t) is the event memory factor at the current time t; M(t-Δt) is the event memory factor at the previous time; α is the forgetting factor; I GSM (t) is the indicator function of the grid support mode; Δf(t) is the deviation between the grid frequency and the reference frequency at the current time t; Δf norm This is the reference value for the frequency deviation used for normalization.

[0038] Preferably, in step S5, the step of calculating the target value of the virtual state charge based on the economic driving factor and the grid driving factor, combined with the dynamically adjusted virtual state charge operating boundary, includes:

[0039] Based on the external environmental data, determine whether a power grid emergency is currently occurring, and perform one of the following steps based on the determination result:

[0040] If it is determined to be a non-grid emergency event, then the economic optimization mode is entered. Under the economic optimization mode, the target value of the original virtual state charge is calculated by combining the economic driving factor and the grid driving factor.

[0041] If it is determined to be a power grid emergency event, then the power grid support mode is entered. Under the power grid support mode, the target value of the original virtual state charge is calculated only based on the power grid driving factor.

[0042] The calculated target value of the original virtual state charge is subjected to amplitude limiting within the dynamically adjusted virtual state charge operating boundary to obtain the final target value of the virtual state charge.

[0043] Preferably, the step of calculating the target value of the original virtual state charge includes:

[0044] Under the aforementioned economic optimization mode, the target value of the original virtual state charge is calculated using the following formula:

[0045] VSoC target_raw (t) = VSoCbase +W·tanh(w eco ·F eco (t)+w grid ·F grid (t));

[0046] In the formula, VSoC base The preset virtual state charge reference value; W is the preset adjustment gain; tanh(·) is the hyperbolic tangent function; w eco and w grid The preset positive weighting coefficient; F eco (t) and F grid (t) represents the economic driving factor and the power grid driving factor at the current moment;

[0047] Under the aforementioned power grid support mode, the target value of the original virtual state charge is calculated using the following formula:

[0048] VSoC target_raw (t) = VSoC base +w ′ grid ·F grid (t);

[0049] In the formula, w ′ grid These are the preset weighting coefficients under the power grid support mode;

[0050] The calculated target value of the original virtual state charge is subjected to a limiting process within the dynamically adjusted virtual state charge operating boundary to obtain the final target value VSoC of the virtual state charge. target (t).

[0051] Preferably, in step S6, the step of calculating the desired charging and discharging power based on the deviation between the target value of the virtual state charge and the current value of the virtual state charge includes:

[0052] The deviation between the target value of the virtual state charge and the current value of the virtual state charge is multiplied by a preset power conversion coefficient to obtain the expected charging and discharging power.

[0053] The deviation between the target value of the virtual state charge and the target value of the virtual state charge is determined, i.e., the control deviation e(t), which is defined as follows:

[0054] e(t) = VSoC target (t)-VSoC(t);

[0055] In the formula, VSoC target(t) represents the target value of the final virtual state charge; VSoC(t) represents the current value of the virtual state charge;

[0056] The formula for calculating the desired charging and discharging power is as follows:

[0057] P exp (t)=K gain e(t) = K gain ·(VSoC target (t)-VSoC(t));

[0058] In the formula, K gain This is the preset power conversion factor.

[0059] This invention provides a control method for user-side energy storage to participate in the operation of a virtual power plant, which has the following beneficial effects:

[0060] 1. This invention sets an event memory factor to record the frequency or severity of historical power grid emergency events, and dynamically adjusts the operating boundary of the virtual state charge based on this. This allows user-side energy storage to proactively reserve more adjustment margin when the power grid fluctuates frequently or is in a high-risk state. This adaptive boundary adjustment mechanism greatly enhances the ability of energy storage units to cope with emergencies within the framework of a virtual power plant, improves the flexibility and resilience of the entire power system, and ensures the reliability of control strategies under complex operating conditions.

[0061] 2. This invention constructs two operating modes: an economic optimization mode and a grid support mode. It intelligently switches between modes based on external environmental data to determine the occurrence of emergencies. Under normal conditions, economic drivers take precedence, enabling peak shaving and valley filling, as well as market arbitrage, for user-side energy storage. However, in the event of a grid emergency, it decisively switches to the grid support mode, prioritizing grid stability. This explicit dual-mode control strategy allows energy storage resources, under the scheduling of a virtual power plant, to balance daily economic benefits with critical grid support responsibilities, achieving a deep exploration and balance of the value of user-side energy storage in grid interaction.

[0062] 3. This invention introduces a virtual state charge, which is not simply a physical quantity of electricity, but a comprehensive indicator integrating expected economic benefits and grid support requirements. The core of the entire control strategy is to perform closed-loop adjustment around the deviation between the current value and the target value of this virtual state charge. This design greatly simplifies the complexity of the control model, providing a unified and clear quantitative control benchmark for user-side energy storage to participate in the operation of virtual power plants, thus simplifying complex coordination control problems. Attached Figure Description

[0063] Figure 1This is a flowchart of a control method for user-side energy storage to participate in the operation of a virtual power plant, according to an embodiment of the present invention.

[0064] Figure 2 This is a schematic diagram of step S4 in an embodiment of the present invention;

[0065] Figure 3 This is a schematic diagram of step S5 in an embodiment of the present invention;

[0066] Figure 4 This is a control system architecture diagram of a user-side energy storage participating in the operation of a virtual power plant, according to an embodiment of the present invention.

[0067] The module includes: 10. Data acquisition and preprocessing module; 20. Driving factor calculation module; 30. Virtual state charge determination module; 40. Event memory and boundary adjustment module; 50. Virtual state charge target value calculation module; 60. Expected charging and discharging power calculation module; and 70. Power correction and command issuance module. Detailed Implementation

[0068] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0069] See Figure 1 and Figure 4 This invention provides a control system for user-side energy storage participating in the operation of a virtual power plant, which also includes a control method for user-side energy storage participating in the operation of a virtual power plant. This control system and method are applied to user-side energy storage devices to adjust their charging and discharging power, thereby balancing the response to virtual power plant dispatching demands with the execution of user-economical operation strategies.

[0070] The control system of this invention may include: a data acquisition and preprocessing module 10, a driving factor calculation module 20, a virtual state charge determination module 30, an event memory and boundary adjustment module 40, a virtual state charge target value calculation module 50, a desired charge and discharge power calculation module 60, and a power correction and command issuance module 70. These modules interact with each other through a preset communication connection (e.g., through a real-time data bus or message queue mechanism within the system) to collaboratively execute the control method of this invention.

[0071] The control method of this invention, in its overall process, may include the following steps:

[0072] Step S1: Obtain real-time operating data and external environment data of user-side energy storage;

[0073] Step S2: Based on real-time operating data and external environment data, determine the economic driving factors and the power grid driving factors;

[0074] Step S3: Determine the current value of the virtual state charge used to characterize the integrated economic and grid support capabilities of user-side energy storage;

[0075] Step S4: Set an event memory factor to record historical power grid emergency events, and dynamically adjust the operating boundary of the virtual state charge based on the event memory factor;

[0076] Step S5: Calculate the target value of the virtual state charge based on the economic driving factor and the grid driving factor, combined with the operating boundary of the dynamically adjusted virtual state charge.

[0077] Step S6: Calculate the expected charging and discharging power based on the deviation between the target value and the current value of the virtual state charge.

[0078] Step S7: Based on the physical constraints of user-side energy storage, the expected charging and discharging power is modified to generate the final power command for responding to virtual power plant dispatch and issue it for execution.

[0079] User-side energy storage refers to the energy storage system installed and used within the power user's premises, mainly used for storing electrical energy and realizing energy management.

[0080] Specifically, the technical solution of this invention achieves control over user-side energy storage by introducing a calculated intermediate variable, namely, virtual state charge. This virtual state charge is a control variable independent of the physical state of charge of the user-side energy storage.

[0081] During the operation of the control system, the data acquisition and preprocessing module 10 executes step S1, the driving factor calculation module 20 executes step S2, the virtual state charge determination module 30 executes step S3, the event memory and boundary adjustment module 40 executes step S4, the virtual state charge target value calculation module 50 executes step S5, the expected charging and discharging power calculation module 60 executes step S6, and the power correction and command issuance module 70 executes step S7.

[0082] See Figure 1 and Figure 4The data acquisition and preprocessing module 10 executes step S1, collecting internal operating data of the user-side energy storage and external market and grid environment data. In a specific embodiment, the data acquisition and preprocessing module 10 connects to the data source through one or more physical communication interfaces. For real-time operating data of the user-side energy storage, the data acquisition and preprocessing module 10 establishes communication with the battery management system (BMS) and power conversion system (PCS) inside the user-side energy storage through an industrial Ethernet interface, using communication protocols such as ModbusTCP or IEC61850. The data acquisition and preprocessing module 10 collects data from the BMS at preset intervals (e.g., 1 second), including physical state of charge (P-SoC), which directly reflects the current actual remaining percentage of battery capacity. It also collects data including the total voltage and total current of the battery cluster; and from the PCS, it collects data including the current active power and reactive power.

[0083] For external environmental data, the data acquisition and preprocessing module 10 connects to the Internet or the dedicated power system data network via a network interface. This module uses an HTTPS-based Application Programming Interface (API) to obtain real-time market electricity prices, such as nodal prices with a 15-minute resolution, from the data publishing platform of the power market operator. Simultaneously, the data acquisition and preprocessing module 10 obtains high-precision grid frequency data from a grid frequency measurement device (e.g., a high-precision power quality analyzer) installed at the user-side energy storage grid connection point, with a data update frequency of no less than 10 times per second.

[0084] To ensure the accuracy of subsequent calculations involved in the virtual power plant's operation and control, the data acquisition and preprocessing module 10 also performs preprocessing operations on the collected real-time operating data and external environment data. All data collected from different sources are appended with a unified timestamp calibrated using the Network Time Protocol (NTP) to ensure data synchronization in time.

[0085] Furthermore, the data acquisition and preprocessing module 10 performs validity checks on the received real-time operating data and external environment data. For example, it checks whether the physical state of charge value is within the valid range of 0 to 100, and whether there are null values ​​or communication interruption flags in the electricity price data. For invalid data, valid data from the previous cycle can be used to replace it, or a preset alarm logic can be activated. After time-stamp synchronization and validity checks, the data is constructed into a unified format data frame by the data acquisition and preprocessing module 10 and transmitted to the control system.

[0086] See appendix Figure 1 and Figure 4 The driving factor calculation module 20 receives data transmitted by the data acquisition and preprocessing module 10, and executes step S2 at the method level to determine the economic driving factor and the power grid driving factor based on real-time operating data and external environment data.

[0087] When determining the driving factor, the driving factor calculation module 20 calculates and determines the economic driving factor F based on the difference between the real-time market electricity price and the preset reference electricity price in the external environment data (for example, if the preset reference electricity price is 0.8 yuan / kWh and the current real-time market electricity price is 0.3 yuan / kWh, then the difference between the two is +0.5 yuan / kWh, and this positive difference will generate an economic driving force to promote energy storage charging). eco (t). The specific calculation method is given by the following formula:

[0088] F eco (t)=k eco ·(C base (t)-C real (t));

[0089] In the formula, F eco (t) represents the economic driving factor at the current time t; k eco This is a preset, dimensionless, positive economic driving coefficient (for example, it can be set to 100 based on the sensitivity requirements of user-side energy storage to electricity price differences; the larger the coefficient value, the greater the economic driving factor value generated per unit price difference); C base (t) represents the preset reference electricity price at the current time t; C real (t) represents the real-time market electricity price at the current time t.

[0090] Simultaneously, the driving factor calculation module 20 calculates and determines the grid driving factor F based on the deviation between the grid frequency and the reference frequency in the external environmental data (e.g., if the reference frequency is 50Hz and the real-time grid frequency is 49.95Hz, then the deviation is -0.05Hz) and the rate of change of the deviation between the grid frequency and the reference frequency (e.g., if the frequency deviation at the current moment is -0.08Hz, and the frequency deviation at the previous moment 1 second ago was -0.05Hz, then the rate of change of this deviation is calculated to be -0.03Hz / s, and this negative rate of change indicates that the grid frequency is accelerating its deviation from the reference). grid (t). This factor is used to quantify the grid frequency deviation required for user-side energy storage to support the operation of virtual power plants. The specific calculation method is given by the following formula:

[0091] Grid driving factor F grid The formula for calculating (t) is:

[0092]

[0093] In the formula, F grid (t) represents the power grid driving factor at the current time t; k p The preset proportionality coefficient is a positive value; k d The preset differential coefficient is positive; Δf(t) is the deviation between the grid frequency and the reference frequency at the current time t, i.e., Δf(t) = f real (t)-50, f real (t) represents the real-time power grid frequency; The rate of change of the deviation between the power grid frequency and the reference frequency is calculated as follows:

[0094] After the calculation is completed, the economic driving factor F eco (t) and grid driving factor F grid (t), which is transmitted to the virtual state charge target value calculation module 50.

[0095] See Figure 1 and Figure 4 The virtual state charge determination module 30 executes step S3 at the method level, and its function is to determine the current value of the virtual state charge. This virtual state charge determination module 30 distinguishes between two scenarios in determining the current value of the virtual state charge: the initial moment and subsequent operating cycles.

[0096] It is important to note that at the initial moment of the control system's startup (t=0), the virtual state charge determination module 30 performs a special initialization operation: it directly sets the initial value of the virtual state charge, VSoC(0), to the initial physical state of charge value, PSoC(0), acquired by the data acquisition and preprocessing module 10 at startup. This step ensures that the virtual state and the physical state are perfectly aligned at the start of control, providing a foundation for subsequent stable control.

[0097] In each operating cycle after the initial moment, the virtual state charge determination module 30 iteratively updates the virtual state charge based on the final execution power fed back by the power correction and command issuance module 70 from the previous control cycle. The specific update method is given by the following formula:

[0098]

[0099] In the formula, VSoC(t) is the virtual state charge of the current control cycle t; VSoC(t-Δt) is the virtual state charge of the previous control cycle; P final (t-Δt) represents the final power command issued and executed in the previous control cycle, where the power value is positive during user-side energy storage discharge and negative during charging; Δt is the operating cycle duration of the control system, for example, 2 seconds; E rated Rated energy capacity for user-side energy storage.

[0100] After completing the calculation, the virtual state charge determination module 30 transmits the current value of the virtual state charge VSoC(t) to the desired charge and discharge power calculation module 60.

[0101] See Figure 1 , Figure 2 and Figure 4 The event memory and boundary adjustment module 40 executes step S4 at the method level. Its function is to dynamically and proactively adjust the operating boundary of the virtual state charge by memorizing the occurrence of historical grid emergency events, thereby optimizing the reserve capacity of user-side energy storage to cope with future grid fluctuations and improving its reliability as part of a virtual power plant participating in grid regulation. In a specific embodiment, the method of the event memory and boundary adjustment module 40 includes the following operations:

[0102] Step S41: Set and update the event memory factor.

[0103] Specifically, the event memory and boundary adjustment module 40 first sets an event memory factor M(t), which is used to record the frequency or severity of historical power grid emergency events. This factor can be understood as a memory index that quantifies recent power grid disturbance history; the larger the value, the more unstable the recent power grid, and the higher the probability that user-side energy storage will be called up to provide support. This memory factor is iteratively updated in each control cycle Δt, and its update formula is:

[0104]

[0105] In the formula, M(t) is the event memory factor at the current time t, and its initial value M(0) can be set to 0. M(t-Δt) is the event memory factor at the previous time. α is a preset forgetting factor, whose value ranges from 0 to 1 (for example, it can be set to 0.01). The role of this forgetting factor is to achieve weighted memory of historical events. The smaller the value of α, the longer the memory retention time of historical events; the term (1-α)·M(t-Δt) reflects the natural decay of historical memory; I GSM (t) is the indicator function of the power grid support mode. The value of this indicator function depends on the judgment of the current power grid state by the event memory and boundary adjustment module 40: when the absolute value of the power grid frequency deviation |Δf(t)| exceeds the preset threshold Δf th When the system enters the grid support mode, I GSM The value of Δf(t) is 1; conversely, in the economic optimization mode, it is 0. This ensures that the memory factor accumulates only during emergency events where the power grid experiences significant frequency deviations. Δf(t) represents the deviation of the power grid frequency from the reference frequency at the current time t. norm This is a reference value for the frequency deviation used for normalization (e.g., 0.5 Hz, representing a very serious frequency drop event). This factor quantifies and normalizes the severity of the current emergency, making the greater the frequency deviation, the greater its contribution to the memory factor.

[0106] Step S42: Dynamically adjust the virtual state charge operation boundary based on the event memory factor.

[0107] Specifically, after updating the event memory factor M(t), the event memory and boundary adjustment module 40 will dynamically adjust the operating boundary of the virtual state charge based on this event memory factor. When the frequency or severity of historical grid emergency events represented by the event memory factor increases, the operating boundary of the virtual state of charge is tightened. Correspondingly, when the frequency or severity of historical grid emergency events represented by the event memory factor decreases, the operating boundary of the virtual state charge is relaxed.

[0108] Specifically, the event memory and boundary adjustment module 40 first defines a default, relatively wide virtual state charge operating range [VSoC]. base_min VSoC base_max (e.g., [20%, 80%]). Then, a dynamic reserve margin R(t) is calculated based on the event memory factor M(t). The reserve margin R(t) is a value calculated based on historical event memory, representing the size of the buffer space that needs to be reserved from the upper and lower ends of the virtual state charge to cope with potential future grid emergencies. The reserve margin R(t) is calculated as follows:

[0109] R(t) = K m ·M(t);

[0110] In the formula, K m This is a preset gain coefficient used to adjust the influence of the memory factor on the degree of boundary tightening. Subsequently, this margin R(t) is used to tighten the default boundary. This tightening refers to reducing the allowable operating range of the virtual state charge by raising the lower operating limit and lowering the upper operating limit, forcing it to remain closer to the middle level. Finally, the dynamic virtual state charge operating boundary [VSoC] at the current moment is obtained. op_min (t),VSoC op_max (t)];

[0111] VSoC op_min (t) = VSoC base_min +R(t);

[0112] VSoC op_max (t) = VSoC base_max -R(t);

[0113] To prevent boundary intersections, constraints must also be applied to the calculation results:

[0114]

[0115] For example, if M(t) increases from 0.1 to 0.5 due to recent frequent grid fluctuations, the dynamically reserved margin R(t) will increase accordingly, and the effective operating range of the virtual state charge (e.g., from [20%, 80%]) will be tightened to [30%, 70%). The essence of this tightening operation is to force the user-side energy storage to maintain more reserve capacity during normal times (keeping both charging and discharging space), so that it can provide the required frequency support more powerfully and for longer during the next grid emergency, thereby improving the operational reliability of the entire virtual power plant.

[0116] Correspondingly, the relaxation of the boundary is achieved through the natural decay mechanism of the event memory factor. Specifically, when the power grid is in a stable operating state for a long time without new emergency events, the event memory factor will gradually decrease due to its built-in forgetting characteristic. The decrease in the event memory factor will directly lead to a corresponding reduction in the reserve margin used to tighten the boundary, thereby lowering the lower limit of operation and raising the upper limit of operation. For example, the operating boundary that was originally tightened to [30%, 70%] will gradually expand as the power grid continues to stabilize, and eventually return to a level close to the benchmark of [20%, 80%]. This automatic relaxation mechanism allows energy storage to have greater operational freedom to pursue economic benefits when the power grid is safe.

[0117] After completing the calculation, the event memory and boundary adjustment module 40 will obtain the dynamic virtual state charge running boundary [VSoC]. op_min (t),VSoC op_max (t)], which is transmitted to the virtual state charge target value calculation module 50.

[0118] See Figure 1 , Figure 3 and Figure 4 The virtual state charge target value calculation module 50 receives the economic driving factor F transmitted by the driving factor calculation module 20. eco (t) and grid driving factor F grid (t), and receive the dynamic virtual state charge running boundary pVSoC transmitted by the event memory and boundary adjustment module 40. op_min (t),VSoC op_max (t)]. The virtual state charge target value calculation module 50 executes step S5 at the method level. Its function is to first calculate a target value of an original virtual state charge based on the driving factor, and then constrain it using a dynamic running boundary to generate the final target value of the virtual state charge that will be transferred to the next step.

[0119] Specifically, the calculation method of the virtual state charge target value calculation module 50 includes the following steps:

[0120] Step S51: Calculate the target value of the original virtual state charge.

[0121] First, it is determined whether a power grid emergency is currently occurring based on external environmental data. In one specific embodiment, this determination is made by comparing the absolute value of the real-time power grid frequency deviation Δf(t) with a preset frequency deviation threshold Δf. th (For example, Δf) th This can be achieved by comparing frequencies (which can be set to 0.1Hz).

[0122] If the judgment result is negative, that is, |Δf(t)|≤Δf thThen, the virtual state charge target value calculation module 50 enters the economic optimization mode. In this mode, the control of user-side energy storage is primarily driven by economic factors, while also considering the conventional power support for the virtual power plant. The virtual state charge target value calculation module 50 uses a nonlinear weighting function to weight and sum the economic and grid driving factors to calculate the original virtual state charge target value VSoC. target_raw (t). The specific calculation method can be given by the following formula:

[0123] VSoC target_raw (t) = VSoC base +W·tanh(w eco ·F eco (t)+w grid ·F grid (t));

[0124] In the formula, VSoC base is a preset virtual state charge reference value (e.g., can be set to 50%); W is a preset adjustment gain used to scale the output range after the combined effect of the driving factors; tanh(·) is the hyperbolic tangent function, the introduction of which makes the weighted summation result exhibit nonlinear characteristics; w eco and w grid For the preset positive weighting coefficients, w in this mode eco The value is greater than w grid To reflect the leading role of economic drivers; F eco (t) and F grid (t) represents the economic driving factor and the power grid driving factor at the current moment.

[0125] If the judgment result is yes, that is, |Δf(t)|>Δf th Then, the virtual state charge target value calculation module 50 enters the grid support mode. In this grid support mode, the control of user-side energy storage is dominated by the grid driving factor, prioritizing frequency support to the grid. The target value of the virtual state charge is calculated by setting the weights of the economic driving factor to zero, and calculating the target value of the original virtual state charge solely based on the grid driving factor. The specific calculation method is given by the following formula:

[0126] VSoC target_raw (t) = VSoC base +w ′ grid ·F grid (t);

[0127] In the formula, w ′ grid The weighting coefficients are preset under the power grid support mode, and their values ​​can be independent of w under the economic optimization mode. gridConfigure the system to meet the requirements for user-side energy storage response capabilities during grid emergencies.

[0128] Step S52: Apply dynamic boundary conditions to constrain the target value.

[0129] The target value VSoC of the original virtual state charge was calculated. target_raw After (t), the virtual state charge target value calculation module 50 does not directly output it, but instead performs a boundary constraint step. This boundary constraint step compares the original target value with the input dynamic virtual state charge running boundary [VSoC]. op_min (t),VSoC op_max The values ​​of (t) are compared and clipped to generate the final target value VSoC of the virtual state charge. target (t). The calculation process is as follows:

[0130] VSoC target (t)=

[0131] max(VSoC op_min (t),min(VSoC op_max (t),VSoC target_raw (t)));

[0132] In the formula, VSoC target (t) is the target value of the final virtual state charge obtained after boundary constraints.

[0133] VSoC op_min (t) is the lower boundary of the virtual state charge operation at the current moment.

[0134] VSoC op_max (t) is the upper boundary of the virtual state charge operation at the current moment.

[0135] VSoC target_raw (t) is the target of the original virtual state charge.

[0136] This formula ensures that regardless of the target value of the original virtual state charge calculated in real time, the final target value of the virtual state charge will be constrained within a more prudent and safe dynamic range determined by historical grid events. This enables the user-side energy storage control strategy to be forward-looking, proactively reserving more adjustment margin during periods of frequent grid fluctuations.

[0137] After completing the calculation, the virtual state charge target value calculation module 50 will obtain the final virtual state charge target value VSoC. target (t), which is transmitted to the desired charge / discharge power calculation module 60.

[0138] See Figure 1and Figure 4 The expected charge / discharge power calculation module 60 receives the current value VSoC(t) of the virtual state charge from the virtual state charge determination module 30, and the target value VSoC of the final virtual state charge from the virtual state charge target value calculation module 50. target (t). The expected charge and discharge power calculation module 60 executes step S6 at the method level, and its function is to calculate the expected charge and discharge power based on the deviation between the target value of the virtual state charge and the current value of the virtual state charge.

[0139] Specifically, in step S6, the step of calculating the desired charging and discharging power includes: multiplying the deviation between the target value of the virtual state charge and the current value of the virtual state charge by a preset power conversion coefficient to obtain the desired charging and discharging power.

[0140] In one specific embodiment, the calculation process first determines the difference between the target value and the current value of the virtual state charge, i.e., the control deviation e(t), which is defined as follows:

[0141] e(t) = VSoC target (t)-VSoC(t);

[0142] In the formula, VSoC target VSoC(t) represents the target value of the final virtual state charge; VSoC(t) represents the current value of the virtual state charge.

[0143] Subsequently, the expected charge / discharge power calculation module 60 multiplies this control deviation e(t) by a preset power conversion coefficient K. gain Thus, the desired charge / discharge power P is obtained. exp (t). The specific calculation formula is as follows:

[0144] P exp (t)=K gain e(t) = K gain ·(VSoC target (t)-VSoC(t));

[0145] In the formula, K gain This is a preset power conversion factor with a clearly defined physical meaning, expressed in kW / %. Its function is to directly convert the dimensionless virtual state charge deviation value into an active power value. The set value of this power conversion factor determines the power intensity of the user-side energy storage response to eliminate the deviation. For example, for a rated power of P... rated User-side energy storage can provide K gain Set to P rated / 100 indicates that when the deviation of the virtual state charge reaches 100%, the expected power will reach the rated power value of the system.

[0146] For example, if the rated power of a user-side energy storage is 500kW, then K gain It can be set to 5kW / %. When the calculated virtual state charge target value at a certain moment is 60%, and the current value is 50%, the difference between the two is +10%. At this time, the expected charging and discharging power P exp (t) is calculated as 5kW / %·10% = 50kW. According to the power sign convention (positive for charging, negative for discharging), this positive value represents a desired charging power of 50kW, which drives the user-side energy storage to charge, bringing its current virtual state charge value closer to the target value. This proportional control-based calculation method is simple in structure and responds quickly, ensuring that the output of the desired power is proportional to the magnitude of the control deviation. This effectively drives the virtual state charge to track its dynamic target, enabling the user-side energy storage to stably participate in the operation and scheduling of the virtual power plant.

[0147] After completing the calculation, the desired charge / discharge power calculation module 60 will obtain the desired charge / discharge power value P. exp (t) is transmitted to the power correction and command issuance module 70 for further processing.

[0148] See Figure 1 and Figure 4 The power correction and command issuance module 70 receives the desired charge / discharge power P from the desired charge / discharge power calculation module 60. exp (t), and simultaneously, in order to perform closed-loop verification of the physical state of user-side energy storage, the power correction and command issuance module 70 also obtains the real-time physical state of charge PSoC(t) from the data acquisition and preprocessing module 10. The power correction and command issuance module 70 executes step S7 at the method level, and its function is to correct the expected charging and discharging power based on the actual operating boundary of user-side energy storage, generate the final power command for responding to virtual power plant dispatch, and execute and issue the final power command.

[0149] In a specific embodiment, the power correction process of the power correction and command issuance module 70 takes into account the physical constraints of user-side energy storage. These constraints specifically include the upper and lower limits of the physical state of charge operation of user-side energy storage, as well as the rated power of the energy storage converter.

[0150] First, the power correction and command issuing module 70 sets the desired charge / discharge power P. exp (t) Verify the rated power of the energy storage converter. The rated power P of the energy storage converter (PCS) ratedPower output (P) refers to the maximum active power that the power electronic device can continuously output or input, measured in kW. This is a critical hardware physical boundary that limits the maximum charging or discharging rate that user-side energy storage can perform at any given time; any power command exceeding this rated value may cause the equipment to overload protection, trip, or even be damaged. Therefore, the power correction and command issuance module 70 specifically specifies the desired charging and discharging power P. exp The absolute value of (t) and the rated power P of the energy storage converter rated Compare them. If |P exp (t)|>P rated If the power demand is too high, then a power limiting process will be applied, which means the expected power will be corrected to its maximum allowable value. For example, if the rated power of a user-side energy storage converter is 500kW, and the calculated expected discharge power is 520kW (discharge is considered positive) or the expected charging power is -520kW (charging is considered negative), then the expected value will be corrected to 500kW or -500kW to ensure that the actual power command issued does not exceed the physical limit of the energy storage converter.

[0151] Secondly, the power correction and command delivery module 70 performs a more critical power prediction correction based on the upper and lower limits of the physical state of charge (PSoC) of user-side energy storage. These upper and lower limits (denoted as PSoC) min and PSoC max Operating limits are the boundaries set by battery manufacturers based on the chemical properties of the battery (e.g., lithium-ion batteries) to ensure safety and extend cycle life. For example, a typical setting is PSoC (Personal Segment Cell). min =15% and PSoC max = 95%. Maintaining the physical state of charge within this range effectively prevents the battery from undergoing deep over-discharge (leading to damage to the active materials of the cell) or overcharge (causing safety issues such as thermal runaway). The power correction and command issuance module 70 performs this correction to prevent the desired power command from causing the battery's actual state of charge to exceed these two safety thresholds.

[0152] The power correction and command issuance module 70 uses the physical state of charge PSoC(t) provided by the current time data acquisition and preprocessing module 10 and the expected power P′ after initial correction of the rated power. exp (t)(For clarity, P′ is used here) exp (t) represents the expected power after the first step of rated power verification), and the predicted physical state of charge (PSoC) that the battery will reach at the end of the current control cycle Δt is calculated in advance. proj (t+Δt):

[0153]

[0154] In the formula, Erated The rated energy capacity for user-side energy storage; η is the charge / discharge efficiency coefficient (η during charging). charge During discharge, it is 1 / η discharge By predicting the P-SoC one cycle from now, it is possible to determine in advance whether there is a risk of overcharging or over-discharging.

[0155] Subsequently, the power correction and command delivery module 70 will use the predicted value PSoC proj (t+Δt) and the preset safe operating range of physical state of charge [PSoC] min PSoC max Compare them.

[0156] If PSoC proj (t+Δt)>PSoC max If overcharging is anticipated, the power correction and command issuance module 70 needs to recalculate the maximum allowable charging power (negative value) for the current cycle. This correction aims to precisely limit the predicted state of charge (SOC) to the PSoC. max Maximum allowable charging power P allowed_charge The formula for calculating (t) is:

[0157]

[0158] For example, if the current P-SoC is 90%, PSoC max If the power is 95%, and it will exceed 95% under the predicted power, then a maximum allowable charging power is calculated according to the above formula. This power value will ensure that the P-SoC reaches or is slightly below 95% after Δt, avoiding overcharging.

[0159] If PSoC proj (t+Δt) <PSoC min (If over-discharge is anticipated), then if PSoC proj (t+Δt) <PSoC min If over-discharge is anticipated, the module needs to recalculate the maximum allowable discharge power (positive value) within the current cycle. This correction will accurately limit the predicted state of charge to the PSoC. min The maximum allowable discharge power P allowed_discharge The formula for calculating (t) is:

[0160]

[0161] For example, if the current P-SoC is 20%, PSoC min If the power is 15%, and it will be lower than 15% under the predicted power, then a maximum allowable discharge power is calculated according to the above formula. This power value will ensure that the P-SoC reaches or is slightly higher than 15% after Δt, avoiding over-discharge.

[0162] Finally, the power correction and command issuance module 70 integrates all the above constraints to generate the final execution power P for responding to the virtual power plant dispatch. final (t). This final power value is the initial expected charge / discharge power P. exp (t), rated power P rated And the permissible power limit (P) calculated based on P-SoC allowed_charge (t) or P allowed_discharge Among the three (t), the most conservative value that can simultaneously satisfy all constraints (i.e., the smallest absolute value among the desired discharge power and all limits) is chosen. This is based on the determination of the safe and feasible final execution rated power P. final After (t), the power correction and command issuance module 70 issues the power command to the user-side energy storage converter (PCS) via an industrial communication bus (e.g., CAN or Ethernet) for execution, thereby achieving precise control of the user-side energy storage. To form a closed loop for the entire control strategy, the power correction and command issuance module 70 also applies this final execution power P... final The value of (t) is fed back to the driving factor calculation module 20. This feedback value will serve as the iterative input for calculating the virtual state charge VSoC(t+Δt) in the next control cycle (i.e., at time t+Δt), thereby ensuring that the evolution of the virtual state can accurately reflect the actual execution of the physical system. This is crucial for the long-term stability of the entire control system during the operation of the virtual power plant.

[0163] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A control method for user-side energy storage participating in the operation of a virtual power plant, characterized in that, include: S1. Obtain real-time operating data and external environment data of user-side energy storage; S2. Based on the real-time operating data and the external environment data, determine the economic driving factor and the power grid driving factor; S3. Determine the current value of the virtual state charge used to characterize the comprehensive economic and grid support capabilities of the user-side energy storage; S4. Set an event memory factor to record historical power grid emergency events, and dynamically adjust the operating boundary of the virtual state charge based on the event memory factor; S5. Based on the economic driving factor and the power grid driving factor, and in conjunction with the operating boundary of the dynamically adjusted virtual state charge, calculate the target value of the virtual state charge. S6. Based on the deviation between the target value of the virtual state charge and the current value of the virtual state charge, the expected charging and discharging power is calculated. S7. Based on the physical constraints of the user-side energy storage, the desired charging and discharging power is modified to generate a final power command for responding to the virtual power plant dispatch and then issued for execution.

2. The control method for user-side energy storage participating in the operation of a virtual power plant according to claim 1, characterized in that, In step S1, the real-time operating data includes the physical state of charge of user-side energy storage; the external environment data includes real-time market electricity prices and grid frequency. Step S1 further includes: attaching a unified timestamp calibrated by the Network Time Protocol to the real-time running data and the external environment data from different sources, and performing validity verification on the real-time running data and the external environment data.

3. The control method for user-side energy storage participating in the operation of a virtual power plant according to claim 2, characterized in that, In step S2, the step of determining the economic driving factor and the power grid driving factor based on the real-time operating data and external environment data includes: The economic driving factor is calculated and determined based on the difference between the real-time market electricity price and the preset reference electricity price in the external environment data. The grid driving factor is calculated and determined based on the deviation between the grid frequency and the reference frequency in the external environment data, and the rate of change of the deviation between the grid frequency and the reference frequency.

4. The control method for user-side energy storage participating in the operation of a virtual power plant according to claim 3, characterized in that, The economic driving factor F eco The formula for calculating (t) is: F eco (t)=k eco ·(C base (t)-C real (t)); In the formula, F eco (t) represents the economic driving factor at the current time t; k eco C is the economic driving factor. base (t) represents the preset reference electricity price at the current time t; C real (t) represents the real-time market electricity price at the current time t; The power grid driving factor F grid The formula for calculating (t) is: In the formula, F grid (t) represents the power grid driving factor at the current time t; k p k is the proportionality coefficient. d is the differential coefficient; Δf(t) is the deviation between the grid frequency and the reference frequency at the current time t; This represents the rate of change of the deviation between the grid frequency and the reference frequency.

5. The control method for user-side energy storage participating in the operation of a virtual power plant according to claim 1, characterized in that, In step S3, the step of determining the current value of the virtual state charge used to characterize the integrated economic and grid support capabilities of the user-side energy storage includes: At the initial moment of the method execution, the initial physical state of charge of the user-side energy storage is used as the current value of the virtual state charge; In subsequent operating cycles, the current value of the virtual state charge is iteratively updated based on the final power command issued and executed in the previous control cycle.

6. The control method for user-side energy storage participating in the operation of a virtual power plant according to claim 5, characterized in that, In step S4, the step of setting an event memory factor to record historical power grid emergency events and dynamically adjusting the operating boundary of the virtual state charge based on the event memory factor includes: Set an event memory factor to record the frequency or severity of historical power grid emergency events; And based on the event memory factor, the operating boundary of the virtual state charge is dynamically adjusted; The step of dynamically adjusting the operating boundary of the virtual state charge includes: When the frequency or severity of historical power grid emergency events represented by the event memory factor increases, the operating boundary of the virtual state charge is tightened. When the frequency or severity of historical power grid emergency events represented by the event memory factor decreases, the operating boundary of the virtual state charge is relaxed.

7. The control method for user-side energy storage participating in the operation of a virtual power plant according to claim 6, characterized in that, The update formula for the event memory factor M(t) is: In the formula, M(t) is the event memory factor at the current time t; M(t-Δt) is the event memory factor at the previous time; α is the forgetting factor; I GSM (t) is the indicator function of the grid support mode; Δf(t) is the deviation between the grid frequency and the reference frequency at the current time t; Δf norm This is the reference value for the frequency deviation used for normalization.

8. The control method for user-side energy storage participating in the operation of a virtual power plant according to claim 7, characterized in that, In step S5, the step of calculating the target value of the virtual state charge based on the economic driving factor and the grid driving factor, combined with the dynamically adjusted virtual state charge operating boundary, includes: Based on the external environmental data, determine whether a power grid emergency is currently occurring, and perform one of the following steps based on the determination result: If it is determined to be a non-grid emergency event, then the economic optimization mode is entered. Under the economic optimization mode, the target value of the original virtual state charge is calculated by combining the economic driving factor and the grid driving factor. If it is determined to be a power grid emergency event, then the power grid support mode is entered. Under the power grid support mode, the target value of the original virtual state charge is calculated only based on the power grid driving factor. The calculated target value of the original virtual state charge is subjected to amplitude limiting within the dynamically adjusted virtual state charge operating boundary to obtain the final target value of the virtual state charge.

9. A control method for user-side energy storage participating in the operation of a virtual power plant according to claim 8, characterized in that, The step of calculating the target value of the original virtual state charge includes: Under the aforementioned economic optimization mode, the target value of the original virtual state charge is calculated using the following formula: VSoC target_raw (t)=VSoC base +W·tanh(w eco ·F eco (t)+w grid ·F grid (t); In the formula, VSoC base The preset virtual state charge reference value; W is the preset adjustment gain; tanh(·) is the hyperbolic tangent function; w eco and w grid The preset positive value weighting coefficient; F eco (t) and F grid (t) represents the economic driving factor and the power grid driving factor at the current moment; Under the aforementioned power grid support mode, the target value of the original virtual state charge is calculated using the following formula: VSoC target_raw (t)=VSoC base +w ′ grid ·F grid (t); In the formula, w ′ grid These are the preset weighting coefficients under the power grid support mode; The calculated target value of the original virtual state charge is subjected to a limiting process within the dynamically adjusted virtual state charge operating boundary to obtain the final target value VSoC of the virtual state charge. target (t).

10. A control method for user-side energy storage participating in the operation of a virtual power plant according to claim 1, characterized in that, In step S6, the step of calculating the desired charging and discharging power based on the deviation between the target value of the virtual state charge and the current value of the virtual state charge includes: The deviation between the target value of the virtual state charge and the current value of the virtual state charge is multiplied by a preset power conversion coefficient to obtain the expected charging and discharging power. The deviation between the target value of the virtual state charge and the current value of the virtual state charge is determined, i.e., the control deviation e(t), which is defined as follows: e(t)=VSoC target (t)-VSoC(t); In the formula, VSoC target (t) represents the target value of the final virtual state charge; VSoC(t) represents the current value of the virtual state charge; The formula for calculating the desired charging and discharging power is as follows: P exp (t)=K gain ·e(t)=K gain ·(VSoC target (t)-VSoC(t)); In the formula, K gain This is the preset power conversion factor.

Images